Merge pull request #2845 from matrix-org/rav/urlcache_error_handling

Handle url_previews with no content-type

synapse/rest/media/v1/preview_url_resource.py

@@ -12,6 +12,19 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+import cgi
+import datetime
+import errno
+import fnmatch
+import itertools
+import logging
+import os
+import re
+import shutil
+import sys
+import traceback
+import ujson as json
+import urlparse
 
 from twisted.web.server import NOT_DONE_YET
 from twisted.internet import defer
@@ -33,18 +46,6 @@ from synapse.http.server import (
 from synapse.util.async import ObservableDeferred
 from synapse.util.stringutils import is_ascii
 
-import os
-import re
-import fnmatch
-import cgi
-import ujson as json
-import urlparse
-import itertools
-import datetime
-import errno
-import shutil
-
-import logging
 logger = logging.getLogger(__name__)
 
 
@@ -286,17 +287,28 @@ class PreviewUrlResource(Resource):
             url_cache=True,
         )
 
-        try:
-            with self.media_storage.store_into_file(file_info) as (f, fname, finish):
+        with self.media_storage.store_into_file(file_info) as (f, fname, finish):
+            try:
                 logger.debug("Trying to get url '%s'" % url)
                 length, headers, uri, code = yield self.client.get_file(
                     url, output_stream=f, max_size=self.max_spider_size,
                 )
+            except Exception as e:
                 # FIXME: pass through 404s and other error messages nicely
+                logger.warn("Error downloading %s: %r", url, e)
+                raise SynapseError(
+                    500, "Failed to download content: %s" % (
+                        traceback.format_exception_only(sys.exc_type, e),
+                    ),
+                    Codes.UNKNOWN,
+                )
+            yield finish()
 
-                yield finish()
-
-            media_type = headers["Content-Type"][0]
+        try:
+            if "Content-Type" in headers:
+                media_type = headers["Content-Type"][0]
+            else:
+                media_type = "application/octet-stream"
             time_now_ms = self.clock.time_msec()
 
             content_disposition = headers.get("Content-Disposition", None)
@@ -336,10 +348,11 @@ class PreviewUrlResource(Resource):
             )
 
         except Exception as e:
-            raise SynapseError(
-                500, ("Failed to download content: %s" % e),
-                Codes.UNKNOWN
-            )
+            logger.error("Error handling downloaded %s: %r", url, e)
+            # TODO: we really ought to delete the downloaded file in this
+            # case, since we won't have recorded it in the db, and will
+            # therefore not expire it.
+            raise
 
         defer.returnValue({
             "media_type": media_type,